You can now set your view to include only the immediate parents, children, or leaves of selected members.

Select one or more members in the Available members or the Current set panel. Go to the Show menu and choose the related members you want to include in the view. All other members are hidden from the view.

Update to predictive forecasting service

Predictive forecast capabilities in Planning Analytics Workspace use a common library of capabilities based on IBM Watson.  In this Planning Analytics Workspace release, this library has been updated to support additional capabilities such as outlier detection and correction.  This library update should not significantly affect the accuracy of previous forecast results. However, you may notice changes in the forecasted values themselves.  The new library creates a more rigorous distinction between training and testing in the historical data, so in some cases, your forecast accuracy may benefit from more historical data.

Note:  In extreme cases, a high number of multiple users simultaneously requesting forecast results might result in a small decrease in performance.  This scenario is unlikely, but if performance decreases the ‘spinning waiting’ icon is displayed until the forecast process is complete.

Note that you can adjust outliers only after you create a preview.

If outliers are detected in your preview, you receive notification.

Click Adjust outliers to update the forecast and see the impact of the adjustment.

An informational message shows the impact of the adjustment. In this example, the adjustment results in a 71% increase in the accuracy of the forecast. You can toggle the Outliers adjusted option  to see the forecast with or without the adjustment. If you choose not to accept the adjustment, you can turn off the Outliers adjusted option and your forecast will not include the automatic outlier adjustment.

When multiple outliers are detected, they are adjusted sequentially from the beginning of the data set. This means each adjusted outlier value is applied to the calculation of the adjustment for the subsequent outlier, resulting in a more accurate forecast.

Changes to cube creation

As of Planning Analytics Workspace 2.0.75, you can no longer create a cube from the Data tree in a book. Cubes can be created only in a Modeling workbench. Additionally, the cube creation process has been modified and an option to create a new cube based on the structure of an existing cube is now available.

Create a new cube on the Modeling workbench

To create a cube:

Create a new cube based on the structure of an existing cube

If a cube exists with the structure and dimensionality that you want you use in a new cube, the Create from cube option provides a handy shortcut for cube creation.

Add variables in the Process editor

Up until now, the number of variables in a Planning Analytics Workspace TurboIntegrator process was limited to the number of columns in the data source. The variable for the process were created automatically, one for each column in the source.

As of Planning Analytics Workspace 2.0.75, you can use the new Add variable option on the Process editor to add variables to a process.

A variable name cannot be empty and must adhere to the Planning Analytics naming conventions. 

You can also delete any variable in a process. Hover the cursor on any variable name to reveal, and then click, the    icon.

The ability to add variables to a process is useful when you are creating a process, but the data source is not available. It’s also useful when the number of columns in the data source changes, as can occur when the data source of a process is modified at run time. 

If you have more variables in a process than columns in the data source, the process still works. The extra variables will simply be empty.
 

Search and filter on hierarchies

The Data tree for books and the Databases tree for workbenches now displays a search box under a dimension when the dimension contains 11 or more hierarchies.

Additionally, the database Search on the modeling workbench now encompasses hierarchies, and you can filter search results to include or exclude hierarchies.

Planning Analytics Workspace 2.0.75 is the last release that lets you create and edit processes in a book. Starting with Planning Analytics Workspace 2.0.76, processes must be created and edited on a modeling workbench.

If you are an administrator or modeler and you attempt to open a process in a book in 2.0.76, you receive a notification that process editing must be performed on a workbench. You can click Open in workbench to immediately open the process on a new workbench.

The 'new experience' is a major update to the Planning Analytics Workspace user experience released as part of the 2.0.56 update in October 2020. The Planning Analytics Workspace new experience provides an updated look and feel, improved navigation, and better content management. It also introduces Applications and Plans, a dedicated modeling environment, and predictive forecasting capabilities.

The user experience in 2.0.55 and earlier releases is referred to as Planning Analytics Workspace Classic. Cloud environments provisioned before the 2.0.56 update could remain on Planning Analytics Workspace Classic, preview the new experience, or permanently switch to the new experience.

The IBM Planning Analytics team intends to remove the capability to remain on Planning Analytics Workspace Classic and preview the new experience during the July 2022 update. Planning Analytics Workspace environments that use the Classic experience will automatically be updated to the new experience as part of the July 2022 update. Planning Analytics cloud customers that use the Classic experience are encouraged to preview and switch to the new experience before July 2022.

If you are still using Planning Analytics Workspace Classic before the July 2022 update, administrators will see a deprecation notice banner on the home page when using Planning Analytics Workspace Classic or previewing the new experience. An administrator can click the More information link to view further details on the deprecation of Planning Analytics Workspace Classic.

This announcement does not impact Planning Analytics on-premises deployments.